NASA Technical Reports Server (NTRS) 20000034070: Laboratory Studies of Low Temperature Rate Coefficients: The Atmospheric Chemistry of the Outer Planets and Titan

13 Pages·0.71 MB·English

by  NASA Technical Reports Server (NTRS)

Final Report
LABORATORY STUDIES OF LOW TEMPERATURE RATE COEFFICIENTS:
THE ATMOSPHERIC CHEMISTRY OF THE OUTER PLANETS AND TITAN
NASA grant No. NAG5-4314
Planetary Atmospheres Program
Technical Officer: Dr. Denis Bogan
Code SR
NASA Headquarters
Washington, D. C. 20546
Grant Period Covered: January 1, 1997 to December 31, 1999
PRINCIPAL INVESTIGATOR:_ Stephen R. Leone
JILA, University of Colorado
Boulder, Colorado 80309-0440
Tel. [phone], Fax x5504
SRL @JILA.COLORADO.EDU
Summary
Laboratory measurements have been carried out with the NASA grant No. NAG5-4314 (January
1, 1997 - December 31, 1999) to determine low temperature chemical rate coefficients of ethynyl
radical (C2H) for the atmospheres of the outer'planets and their satellites. This effort is directly
related to the Cassini mission which will explore Saturn and Titan. A laser-based
photolysis/infrared laser probe setup was used to measure the temperature dependence of kinetic
rate coefficients from =150 to 350 K for C2H radicals with H2, C2H 2, CH 4, CD 4, C2H 4, C2H 6,
C3H8, n-C4H10, i-C4H10, neo-CsHl2, C3H 4 (methylacetylene and allene), HCN, and CH3CN.
The results revealed discrepancies of an order of magnitude or more compared with the low
temperature rate coefficients used in present models. A new Laval nozzle, low Mach number
supersonic expansion kinetics apparatus has been constructed, resulting in the first measurements
of neutral C2H radical kinetics at 90 K and permitting studies on condensable gases with
insufficient vapor pressure at low temperatures. New studies of C2H with acetylene have been
completed.
2
Final Progress Report NAG5-4314
This program involves the measurement of low temperature rate coefficients for neutral
radical reactions that are important in models of the planetary atmospheres of Saturn, Titan, and
Jupiter. A key goal is to obtain benchmark sets of rate coefficients for a number of different
chemistries that will be needed to help analyze the results of the Cassini/Huygens mission; the
results are also immediately applicable to recent Galileo measurements and models for Jupiter.
Research under the three-year grant NAG5-4314, "Laboratory Studies of Low Temperature Rate
Coefficients: The Atmospheric Chemistry of the Outer Planets," provided an outstanding number
of new measurements of kinetic rate coefficients for CzH radical reactions with many of the
relevant saturated and unsaturated hydrocarbon molecules. Using a laser pump/probe apparatus,
most of these measurements were made in a low temperature flow cell over the temperature range
from 300-360 K down to 150-170 K. Through a major equipment grant, a conceptually different
Laval nozzle apparatus was designed and constructed during the previous grant period to achieve
much lower temperatures, which is especially important to study condensable gases, i.e. gases
with insufficient vapor pressures at low temperatures. The first measurements of rate constants for
hydrocarbon radical reactions down to 90 K were achieved with this apparatus. Below is a
description of the two experimental techniques used and the recent results and their interpretations.
Laser Pump/Probe Results of Low Temperature Kinetics
The laser pump/probe apparatus consists of an ultraviolet laser for photolysis and an
infrared probe laser for transient absorption of the C2H radicals. The rate measurements are
carried out as a function of temperature in a transverse flow cell designed specifically for achieving
low temperatures.
In a typical experiment, a 193 nm argon fluoride pulsed excimer laser is used to photolyze a
suitable precursor molecule (acetylene, C2H2) to produce C2H, and a high resolution, tunable
infrared F-center laser probes the transient concentrations of the radical species directly in
absorption to extract the kinetic rate coefficients. The experiment involves flowing a mixture of a
precursor molecule, such as C2H 2and a reactant, such as H2, methane (CH4), acetylene, or larger
hydrocarbons in a rare gas buffer through the I00 cm long cell. The photolysis beam and probe
laser beam (multipassed to increase the absorbance) are overlapped over the length of the cell by
means of dichroic mirrors which reflect one wavelength and are transparent to the other. The cell
is operated with various cooling fluids and a low temperature circulating pump to achieve
temperatures routinely down to 150 K. The kinetics are carried out in the pseudo-first order
regime so that the observed time decays of the radical signals are related to the first order removal
rates of the radical species.
In Table 1below, our measured rate coefficients for the C2H + hydrocarbon reactions and
cyanide compounds are summarized. The measurements are made over a range of temperatures,
but only the lowest temperatures are shown in the table. New measurements on the larger
hydrocarbon species (C3H 8, i-C4H10, n-CnHIo, neo-CsHl2, CH3CCH, and CH2CCH2)and the
coupling reactions with nitrogen compounds (HCN, CH3CN) were the focus of the experiments in
the last three years. In the table, the measurements are compared with those values currently used
in many of the atmospheric models (Gladstone et al. 1996). In addition to the Gladstone et al.
3
(1996) model for the atmosphere of Jupiter, other recent major models for Titan's atmosphere are
given by Toublanc et aI. (1995) and Lara et al. (1996). These add to the already existing models of
Yung et al. (1984, 1987) developed for Titan. While there are some differences in the rate
coefficient values used in these models, a detailed assessment of the rate coefficients used in each
model is not considered here, but discussions are included in each of our publications on the
specific reactions.
As can be seen from the table, there are considerable discrepancies (30% to an order of
magnitude) between the accurately determined experimental values and many of the quantities
presently used in the models. In some cases the models are not that sensitive to a particular rate
and the use of a slightly different number is not a problem. In others, the models use only room
temperature values. In still others, the use of the new determinations would cause the
concentrations of some species to be altered even further from expectations. For example, in
Gladstone et al. (1996) the rate coefficients of key C2H reactions are adjusted to maximize the
ethane to acetylene ratio for Jupiter using only older literature values, rather than the newer
determinations, which would make the agreement even worse. This practice could result in
Table 1 Some low temperature rate coefficients measured in this work compared
to values presently used in models, a)
Reaction Temperature Measured (cm 3s"l) Gladstone et al. 1996
C2H + H2 178 K 5.0 x 1014 1.7 x 1014
C2H + CH 4 154 K 5.5 X 10 -13 1.8 X 10-12
C2H + C2H2 143 K 1.9 x 10-l° 1.5 x I01°
153 K 3.3 x 10ll 2.1 x 1011
C2H + C2H6
150 K 1.4 x 10l° 2.0 x 1011
C_H + C2H4
180 K 1.1 x 101° 1.4 x 10-11
C2H + C_H8
177 K 1.0 x 10-1° 1.4 x 1011
C2H + i-C4HI0
176 K 1.6 x 101° 1.4 x 10-11
C2H + n-C4HIo
181 K 1.3 x 10 "10 none
C_H + neo-C_HI2
155 K 2.4 x 101° none
C_H + CH3CCH
159 K 2.4 x 10!° none
C_H + CH2CCH _
297 K 3.9 x 10-13 2.2 x 1012
C2H + HCN
C H+CH CN 262 K 1.0 x 1012 none
2 3
a)No error bars are shown, see publications. Most measurements are carried out over a wide
temperature range, from the lowest value indicated in the table to 300 K or 360 K, and a
complete temperature fit is formulated for each reaction. Each measurement consists of many
data runs.
important pathways being overlooked. In addition, sometimes the models may have relied too
4
heavily on the older Voyager data for species concentrations, resulting in inappropriate adjustments
to the model parameters.
Summary of Results
The results for the reactions of C2H with H2 (Opansky and Leone 1996a), CH 4 (Opansky
and Leone 1996b), CEH 2(Pedersen et al. 1993, Opansky and Leone 1996b), C2H 4 (Opansky and
Leone 1996a), C2H 6 (Opansky and Leone 1996a), and 02 (Opansky et al. 1993) were discussed
in detail in the previous proposal. We briefly discuss below the interpretation of some of the more
recent results in the table above.
C2H + Calls, n-C4Hlo, i-C4Hlo , and neo-CsH12
The reactions of C2H with large saturated and unsaturated hydrocarbons are not well
characterized. Gladstone et at. (1996) urged that there be more rate determinations of radicals with
C3 and larger hydrocarbon species, and they indicated that these rates will be crucial to understand
the carbon partitioning on Jupiter. One of the difficulties with studies of these species is that the
gases start to be condensable at low temperatures. Our results for C2H with C3H 8, n-C4Hlo, i-
C4HI0, and neo-C5H12 (Hoobler et al. 1997) are the only absolute rate coefficient determinations
ever reported for these processes and the measurements are successfully made down to 150-180 K
in the flow ceil, depending on species. The measurements show that the C2H reactions with the
larger hydrocarbons have very rapid rates, much larger than the estimates used in the models
(Table 1). In some models these reactions are not yet included.
The results for propane are nearly independent of temperature, whereas for n-butane,
isobutane, and neopentane the temperature dependencies are negative (Hoobler et al. 1997),
indicating that the rates are increasing with decreasing temperature. The negative temperature
dependencies stress the need for the Laval nozzle measurements, which will be carried out at even
lower temperatures and are discussed later. The values presently used in the models of the
planetary atmospheres are in error by nearly an order of magnitude, even before extrapolation to
lower temperatures, and at the lowest temperatures of 70 K on Titan, the actual rate coefficients
may differ even more. For these reactions, the role of possible long-lived collision complexes
must be recognized in the dynamical mechanisms of the reactions. These reactions are essential to
the understanding of the synthesis of larger molecules in the planetary atmospheres, since they
create larger radicals which can then recombine, affecting the ratios of various hydrocarbons in the
atmosphere. Thus the formation of adduct species, which could be important products of these
reactions, will also need to be determined.
C2H + CH3CCH and CH2CCH 2
The reactions of C2H with methylacetylene (CH3CCH) and allene (CH2CCH2) were
identified by Gladstone et al. (1996) as key processes involving Ca species that need to be included
in the atmospheric models, but which have not been incorporated to date. Reactions with other
species that have similar concentrations, however, such as diacetylene, cyanoacetylene, and
cyanogen, have already been included. Our results are the first determinations of reaction rate
coefficients for C2H with C3H 4 molecules (Hoobler and Leone 1999). Both reactions are fast, and
thetemperaturedependencieasreeitherflatorslightlynegativep,redictingsignificantlyfaster
reactionsasthetemperaturebecomeslower.Atthelowesttemperaturesstudied,155K,
condensationoftheC3H4speciesbecomesaproblem,andthusanobservedlevelingoff oftherate
constantsbelow200K ismostlikely duetocondensationlossofthereactants.Therate
coefficientsundertheconditionsoftheplanetaryatmosphereasrelikelytobeevengreater.
Thesereactionshavemanypossibleproducts.Forexample,thereactionofC2Hwith
methylacetylenecanformfive possibleproducts:C2H2+CH2CCH,C2H2+CH3CC,C4H2+
CHy C5H 4 + H, and C5H 5. Some of the products involve abstraction processes and others
involve adduct formation. In future work it will be important to determine which of the various
product species are formed and their branching ratios. Although some aspects of the
photochemistry of methylacetylene and allene have been included in the models of the planetary
atmospheres of Titan and Jupiter, these processes have often emphasized only photolysis,
reactions with H atoms, and condensation. As the models become more complete, the reactions of
C3H 4 species with C2H can now be included with accurate rate coefficients as well.
C2H + HCN and CH3CN
Studies of the reaction of C2H with HCN and CH3CN are particularly troublesome because
of the ease of condensation (low vapor pressure) of these nitrogen-containing compounds at low
temperatures. Thus our measurements (Hoobler and Leone 1997) with the flow cell apparatus
were limited to a very narrow temperature range of 262-360 K. In addition, while some estimates
suggested that these reactions might be rapid, in fact we find the reactions are extremely slow, have
a significant barrier, and are therefore even more difficult to measure. Our own ab initio
calculations show that the lowest-barrier transition state of the C2H + HCN reaction leads to HCaN
+ H products, formed via an addition-elimination mechanism; the calculations confirm that there is
a high barrier, which is the reason that the reaction is slow.
In the current atmospheric models of Titan (Lara et aL 1996), the rate coefficient assumed
for the reaction of C2H + HCN _ HC3N + H and the eddy diffusion coefficient are critical
features in obtaining the abundance of HC3N. The much smaller value of the rate coefficient
obtained here compared to that used by Lara et al. (more than a factor of 30 for the room
temperature value and 200 times smaller when the experimental temperature dependence is
extrapolated to 170 K) has significant impact on the modeled HC3N distribution and the
discrepancies with Voyager/IRIS measurements. The model (Lara et al. 1996) predicts vertical
profiles for the HC3N mixing ratio that are too high by nearly two orders of magnitude. Both
reactions decrease substantially with decreasing temperature, suggesting that at the lowest
temperatures on Titan, these reactions will be very slow indeed.
Other models utilize different eddy diffusion coefficients, photolysis production rates, and
assumptions about the neutral radical rate coefficients. Therefore it is not straightforward to
compare resulting numbers from each model without detailed knowledge of the contributing and
canceling factors. A detailed analysis of the parameters for several models is contained in Hoobler
and Leone (1997). However, it is valuable to present the results of a reassessment of the
predictions made by Hoobler and Leone (1997) for the HC3N production rates. This reassessment
is based on just the changes in two experimental rate coefficients, for the C2H + HCN reaction
measuredhereandtheCN+C2H 2reaction measured by Sims et al. (1993). The results in Table 2
reveal how dramatic the affect is on the calculated HCaN species.
Table 2 Reassessment of the Production Rates for HC3N Species Compared to
Models. a)
Reaction Path Reassessment Yung et al. 1984 Toublanc et al. Lara et al. 1996
1995
C2H + HCN -4 0.008 b) 0.3 0.5 0.3
HC3N + H
CN + C2H2-4 14c) 2 75 3600
HCaN + H
a)In molecules cm "3s"I
b)Using k=5 x 10"14 based on extrapolation to 170 K.
e)Using k=4 x 10"10 from Sims et al. 1993.
Laval Nozzle Apparatus for Low Temperature Kinetics
As noted above, many of the reactions of interest for the outer planets occur at temperatures
down to 70 K, for example on Titan. Because of the low vapor pressure of many reagents, the
gases are condensable at low temperatures, and the kinetic determinations of most reactions can
only be measured in flow cells down to 150 K or, in many cases, only at 300 K. As shown in our
studies, there are often dramatic changes in the rate coefficients at lower temperatures, and it is
often not possible to predict whether the rates will increase, decrease, or remain the same upon
lowering the temperature.
An important instrumental development was made in this grant period with the support of a
major equipment grant from NASA (Lee, Hoobler, and Leone 2000). We constructed a low Mach
number supersonic Laval nozzle expansion as a new apparatus for making neutral kinetics
measurements at temperatures much lower than possible with a flow cell. The Laval nozzle
expansion produces a long (20 cm) region of uniform low-temperature gas, which can be used for
accurate kinetics measurements down to 20 K (Sims et al. 1992, Hawley et al. 1990, Smith and
Hawley 1992, Sims et aI. 1993, Sims et al. 1994, Atkinson and Smith 1994, Atkinson and Smith
1995). The Laval nozzle expansion creates the constant low temperature zone over which the
reactions can proceed without the interference of walls. The continuous flow Laval nozzle method
requires an enormous pumping system, which is beyond the limits of reasonable laboratory size
and expense. However, the method of Atkinson and Smith uses a pulsed valve source for the
supersonic Laval nozzle expansion and this method was adapted in our laboratory with rather
modest pumps and cost. We also developed a rather general single photon laser ionization scheme
for detection of radical and neutral species; this method uses the 9th harmonic of the Nd:YAG laser
at 118 nm (10.5 eV per photon), which had already been demonstrated for polyacetylene species
(Bandy et al. 1992, Frost et al. 1995) and used in our laboratory for molecular beam epitaxy
species (Strupp et aI. 1993).
Figure1showsthe Laval nozzle apparatus constructed in our laboratory. It consists of a
pulsed valve and a shaped Laval nozzle to produce a uniform low temperature gas flow, a 193 nm
photolysis laser to generate the radical of interest, a skimmer to sample the gas stream, impact
pressure transducers to characterize the beam, the 118 nm ionization laser, and a time-of-flight
mass spectrometer. For kinetics the nozzles are shaped to produce a uniform density of low
temperature gas over a long path length.
TP
< 10-6 Torr
118 nm
Pulsed valves _ X_
Laval nozzle \ (ionization)
-.,,
193 nm
I ----_ TP
(photolysis)
"'_iKigg""
block MP DP
Nozzle translator 0.1-1 Tort 10-4 Torr
Figure 1 Pulsed Laval nozzle apparatus for low temperature kinetics.
A uniform density of radicals is produced by laser initiation coaxially along the path of the
expansion using a pulsed excimer laser at 193 nm. The pulsed laser ionization probe is used to
detect the products either as a function of distance along the expansion, a function of concentration
of the reactant, or by precision time delays between the two pulses. We have provision to
substitute a resonantly enhanced multiphoton ionization (REMPI) process to probe selectively an
individual radical species of interest. The ions are sampled from the flow with a skimmer and
extracted into the entrance of a differentially pumped time-of-flight mass spectrometer. The laser
ionization time-of-flight mass spectrometer provides a very general method of detection for both
reactants and many product species. The main criterion is that the ionization potential of the
species is lower than 10.5 eV. By addition of reactive reagents into the expansion gas and accurate
calibration of the flow density, the change in radical concentration as a function of distance or time
delay is used to measure the reaction rate constant under pseudo first order conditions. The
reaction distance is readily varied by changing the distance of the probe laser from the nozzle exit,
and the timing between the two lasers sets accurately the time for the reaction to proceed. Using a
variety of nozzles, and with heating and cooling of the nozzle body, expansions can be achieved
with temperatures ranging from 70 K to 300 K.
A Laval nozzle is a converging-diverging channel for obtaining prescribed expansions.
ThefirstpartoftheLavalnozzleisashortconvergenstectionwheretheflowbeginstoaccelerate
fromnearzeronethydrodynamicvelocityinthestagnationregion.Atthethroatwherethenozzle
diameterisaminimum,theflowreachesMach1.Followingthethroat,givenasufficientpressure
dropacrossthenozzle,supersonicflowisachievedinthedivergentregion. Inthenozzledesign,
accomplishedwithcodesprovidedbyM. SmithatArizona,theradiusandslopeatthethroatmust
becarefullymatchedtothebeginningofthedivergentsection,sothatshockwavesdonotoccur.
A slipgasisaddedaroundtheexpansiontoprovideaconstansttagnangtaspressureatthewalls
andcollimatethegasflow.
Figure2showstheresultofimpactpressuretransducemr easurementosfthebeamfrom
oneoftheLavalnozzlesconstructedforthesestudies.Thisparticularnozzlewasdesignedto
operateatMach3,andit wasfoundtoproduceMach3.35withauniformtemperature in the gas
flow of 89+4 K at a pressure of 0.2 Torr. In the figure the excellent uniformity of the expansion
and minimal spreading of the beam can be seen.
Impact Pressure Map of Mach 3.4 N2Flow
9 f 0 2 4 6 $ 10 12
radial 13.5
distance t axial distance /em
mm
Fig. 2 Impact pressure transducer measurements of the Laval nozzle output
as a function of distance downstream of the flow.
A typical experiment is performed in the following manner. The photolytic precursor gas
for the radical of interest, the desired reactant gas, and a cartier gas are admitted by the pulsed
valve for a 5 ms duration. The supersonic expansion forms and becomes uniform. At an
appropriate time after the uniform expansion characteristics have been achieved, the pulsed excimer
laser is puIsed coaxially along the expansion to prepare a uniform radical density by photolysis of
the precursor. The radicals begin to react with the desired reagent. At a measured time delay and
given distance in the flow stream after the photolysis pulse, the probe laser is pulsed to interrogate
the remaining radical density by producing ions. The integrated ion signals are collected by the
TOF-MS and measured with the data acquisition. Measurements can be taken either at various
distances in the expansion at a fixed time delay or at various time delays at a fixed distance to
9
extracttherateinformation,asafunctionofreactanctoncentrationI.mpactpressuretransducers
(Fig.2above)areusedtocharacterizethenozzleexpansionMachnumberand,fromthat
determinationt,hedensityandthenthetemperaturaerecomputed(AtkinsonandSmith1995).
ThesetemperaturecsanbecheckedbymeasuringarotationalpopulationdistributionbyREMPIfor
atypicalgaslikeNO. Undernonreactiveconditionst,hesignalsfromtheprobelasercanalsobe
usedtodeterminetheuniformityoftheradicaldensityalongtheexpansionlength.The
temperatureisvariedbyusingavarietyofremovablenozzlesandbyheatingandcoolingthenozzle
bodywithacirculatingfluid.
IntheexperimentsthegaspulsefromtheLavalnozzleis5mslong,andthe20nsexcimer
laserpulseirradiatesonlyaverysmallslugofgasalongthelengthoftheentiregaspulse.InFig.
3,timingdiagramdepictsanactualgaspulse,thesmallslugofgasirradiatedbytheexcimerlaser
pulseatatimeof3ms,andanexpandedviewoftheportionofthegaspulsethatisphotolyzed,
_ psflow
r ; ;
i
193 nm --o- eidxpecaril_l
C4H 6_ C3FI3+Ctt 3
c4a5+H [c4nd _,
2.9 3 3.1 3.2 3.3 3.4
time / ms
Figure 3 Timing diagram with measured data for the Laval nozzle gas pulse
and the excimer photolysis pulse, detected by the 118 nm laser.
10